Geometric error identification of rotary axes based on uniaxial motion measurement of double ball bar

In order to avoid the negative influence of the interference sources such as the geometric errors and servo-control errors of non-identified axes on the measurement results, and lower the complexity of the identification model, this paper proposes an identification method for the geometric errors of rotary axes based on uniaxial motion measurement of double ball bar (DBB), which uses uniaxial motion mode to replace the conventional multiaxial linkage measurement mode. Firstly, the mathematical models of the bar length change (ΔL) under the influence of the position-dependent geometric errors of the rotary worktable is established based on homogeneous coordinate transformation theory. Then, through analyzing the installation parameters of the DBB, an error identification model is constructed based on the column full-rank identification matrix, based on which an identification scheme including nine independent measurement tests is designed. The scheme has high robustness through increasing the number of the measurement tests, which reduces the possible identification accuracy loss caused by the relatively large measurement errors in a single test. In the identification experiments, accurate installation of the tool ball was achieved through iterative adjustment, and the influences of the installation errors of the workpiece ball and the position-independent geometric errors of the worktable were excluded. Finally, the prediction analysis of ΔL and the experiment of error compensation were performed. The maximum absolute value of ΔL in one rotation of the worktable is reduced from 0.010 3 mm to 0.002 0 mm after compensation, which verifies the effectiveness of the identification method. © 2020, Science Press. All right reserved.